Speech perception in adverse listening environments (e.g., noise) is more difficult than in quiet even for individuals with normal hearing. The goal of the proposed research program is to evaluate a novel assessment and training protocol for individuals who experience difficulty understanding speech in noise. The two target populations are non-native (L2) English speakers and native English speakers with various forms of auditory processing disorders (referred to generically as APD). Similar problems in noise have been reported for both groups, but previous studies have suggested several different underlying mechanisms. For L2 speakers, it has been assumed that their perceptual problems in noise are due primarily to acoustic-phonetic interference from their native language worsened by masking from noise. That is, their problem is a diminished ability to use context-independent (bottom-up) processing. To compensate, this group tends to rely heavily on the contextual information (top-down processing). The underlying mechanisms for APD group are not clearly understood. While some argue that difficulties in noise are caused by impaired auditory temporal coding, others have suggested that this group may also have poor acoustic-phonetic awareness. Although models of speech perception differ in terms of how these skills work in combination in normal native speakers, there seems to be agreement that bottom-up processing is automatic, whereas top-down processing can be more effortful and is influenced by the integrity of bottom-up processing. Thus the overuse of top-down processing to compensate for poor bottom-up processing may impede overall efficiency of speech perception. The proposed research program will test the following three hypotheses: 1) that the difficulties perceiving speech in noise for both L2 speakers and native APD groups are related to compromised acoustic-phonetic awareness (i.e., bottom-up processing);2) that an improvement in bottom-up processing skills will aid speech perception in noise by reducing the need to rely as heavily on contextual information (i.e., top-down processing);and 3) that explicit training in bottom-up processing will improve speech perception in noise for both groups. These hypotheses will be tested through three aims. In Aim 1, listeners'use of top-down and bottom-up processing to understand speech in noise will be investigated by comparing perception of isolated words and words in sentences. In Aim 2, a novel bottom-up processing training procedure using sentences with little contextual information will be evaluated with a subgroup of L2 listeners selected under Aim 1. In Aim 3, this training procedure will be evaluated with a subgroup of APD listeners. The proposed research is an important initial stage in establishing an effective and efficient assessment and training protocol for individuals with APD and for L2 speakers of English. It is also expected to provide important new information on the underlying perceptual mechanisms used by individuals who experience difficulty understanding speech in noise. Relevance: The ultimate goal of this research project is to develop a protocol for the assessment and training of speech perception for individuals who experience marked difficulty understanding speech in noise. The efficacy of a novel auditory training protocol will be evaluated in two groups of individuals who are known to have difficulty understanding speech in noise: English speaking children with auditory processing disorders and/or language-based learning disabilities and adult L2 speakers of English.